Md Mushfiqur Rahman commited on
Commit
f45dc62
1 Parent(s): 988449c

Upload with huggingface_hub

Browse files
README.md ADDED
@@ -0,0 +1,15 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: apache-2.0
3
+ tags:
4
+ - canine
5
+ - pretrained-on-english-language
6
+ ---
7
+
8
+ ### How to use
9
+
10
+ Here is how to use this model:
11
+
12
+ ```python
13
+ from transformers import CanineModel
14
+ model = CanineModel.from_pretrained('mushfiqur11/<repo name>')
15
+ ```
all_results.json ADDED
@@ -0,0 +1,22 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 78.95,
3
+ "eval_las": 73.77547916032857,
4
+ "eval_loss": 5.533295154571533,
5
+ "eval_runtime": 6.408,
6
+ "eval_samples": 909,
7
+ "eval_samples_per_second": 141.853,
8
+ "eval_steps_per_second": 17.79,
9
+ "eval_uas": 80.46175168170909,
10
+ "test_las": 75.52606021366138,
11
+ "test_loss": 4.81597900390625,
12
+ "test_runtime": 4.9664,
13
+ "test_samples": 680,
14
+ "test_samples_per_second": 136.92,
15
+ "test_steps_per_second": 17.115,
16
+ "test_uas": 82.21107154418907,
17
+ "train_loss": 0.1834182323773702,
18
+ "train_runtime": 7377.9639,
19
+ "train_samples": 6075,
20
+ "train_samples_per_second": 65.059,
21
+ "train_steps_per_second": 2.033
22
+ }
config.json ADDED
@@ -0,0 +1,109 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "google/canine-s",
3
+ "architectures": [
4
+ "CanineForBiaffineParsing"
5
+ ],
6
+ "attention_probs_dropout_prob": 0.1,
7
+ "bos_token_id": 57344,
8
+ "downsampling_rate": 4,
9
+ "eos_token_id": 57345,
10
+ "hidden_act": "gelu",
11
+ "hidden_dropout_prob": 0.1,
12
+ "hidden_size": 768,
13
+ "id2label": {
14
+ "0": "_",
15
+ "1": "acl",
16
+ "2": "advcl",
17
+ "3": "advmod",
18
+ "4": "amod",
19
+ "5": "appos",
20
+ "6": "aux",
21
+ "7": "case",
22
+ "8": "cc",
23
+ "9": "ccomp",
24
+ "10": "clf",
25
+ "11": "compound",
26
+ "12": "conj",
27
+ "13": "cop",
28
+ "14": "csubj",
29
+ "15": "dep",
30
+ "16": "det",
31
+ "17": "discourse",
32
+ "18": "dislocated",
33
+ "19": "expl",
34
+ "20": "fixed",
35
+ "21": "flat",
36
+ "22": "goeswith",
37
+ "23": "iobj",
38
+ "24": "list",
39
+ "25": "mark",
40
+ "26": "nmod",
41
+ "27": "nsubj",
42
+ "28": "nummod",
43
+ "29": "obj",
44
+ "30": "obl",
45
+ "31": "orphan",
46
+ "32": "parataxis",
47
+ "33": "punct",
48
+ "34": "reparandum",
49
+ "35": "root",
50
+ "36": "vocative",
51
+ "37": "xcomp"
52
+ },
53
+ "initializer_range": 0.02,
54
+ "intermediate_size": 3072,
55
+ "label2id": {
56
+ "_": 0,
57
+ "acl": 1,
58
+ "advcl": 2,
59
+ "advmod": 3,
60
+ "amod": 4,
61
+ "appos": 5,
62
+ "aux": 6,
63
+ "case": 7,
64
+ "cc": 8,
65
+ "ccomp": 9,
66
+ "clf": 10,
67
+ "compound": 11,
68
+ "conj": 12,
69
+ "cop": 13,
70
+ "csubj": 14,
71
+ "dep": 15,
72
+ "det": 16,
73
+ "discourse": 17,
74
+ "dislocated": 18,
75
+ "expl": 19,
76
+ "fixed": 20,
77
+ "flat": 21,
78
+ "goeswith": 22,
79
+ "iobj": 23,
80
+ "list": 24,
81
+ "mark": 25,
82
+ "nmod": 26,
83
+ "nsubj": 27,
84
+ "nummod": 28,
85
+ "obj": 29,
86
+ "obl": 30,
87
+ "orphan": 31,
88
+ "parataxis": 32,
89
+ "punct": 33,
90
+ "reparandum": 34,
91
+ "root": 35,
92
+ "vocative": 36,
93
+ "xcomp": 37
94
+ },
95
+ "layer_norm_eps": 1e-12,
96
+ "local_transformer_stride": 128,
97
+ "max_position_embeddings": 16384,
98
+ "model_type": "canine",
99
+ "num_attention_heads": 12,
100
+ "num_hash_buckets": 16384,
101
+ "num_hash_functions": 8,
102
+ "num_hidden_layers": 12,
103
+ "pad_token_id": -100,
104
+ "torch_dtype": "float32",
105
+ "transformers_version": "4.17.0",
106
+ "type_vocab_size": 16,
107
+ "upsampling_kernel_size": 4,
108
+ "use_cache": true
109
+ }
eval_predictions.csv ADDED
The diff for this file is too large to render. See raw diff
 
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:51761b23310641c535a3d9cf016a19130e78f3279bf8335810299a28320de5bd
3
+ size 620816369
special_tokens_map.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"bos_token": {"content": "", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}, "eos_token": {"content": "", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}, "sep_token": {"content": "", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}, "pad_token": {"content": "\u0000", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}, "cls_token": {"content": "", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}, "mask_token": {"content": "", "single_word": false, "lstrip": true, "rstrip": false, "normalized": true}}
test_predictions.csv ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer_config.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"bos_token": {"content": "", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true, "__type": "AddedToken"}, "eos_token": {"content": "", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true, "__type": "AddedToken"}, "sep_token": {"content": "", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true, "__type": "AddedToken"}, "cls_token": {"content": "", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true, "__type": "AddedToken"}, "pad_token": {"content": "\u0000", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true, "__type": "AddedToken"}, "mask_token": {"content": "", "single_word": false, "lstrip": true, "rstrip": false, "normalized": true, "__type": "AddedToken"}, "add_prefix_space": false, "model_max_length": 2048, "special_tokens_map_file": "/home/mrahma45/.cache/huggingface/transformers/00ffef2919c64d09945b6c06b1c03c4d18bf55f88013390d4443a1eb8c4f8a10.ab71f530366fe02e2834427e7b90198bfd0d573bc4279bfafdb2b95fe2b46dde", "name_or_path": "google/canine-s", "tokenizer_class": "CanineTokenizer"}
trainer_state.json ADDED
@@ -0,0 +1,1225 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 73.77547916032857,
3
+ "best_model_checkpoint": "/scratch/mrahma45/pixel/finetuned_models/canine/canine-base-finetuned-parsing-ud-Arabic-PADT/checkpoint-15000",
4
+ "epoch": 78.94736842105263,
5
+ "global_step": 15000,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.53,
12
+ "learning_rate": 7.6e-05,
13
+ "loss": 4.5897,
14
+ "step": 100
15
+ },
16
+ {
17
+ "epoch": 1.05,
18
+ "learning_rate": 7.948993288590604e-05,
19
+ "loss": 2.0458,
20
+ "step": 200
21
+ },
22
+ {
23
+ "epoch": 1.58,
24
+ "learning_rate": 7.895302013422819e-05,
25
+ "loss": 1.5969,
26
+ "step": 300
27
+ },
28
+ {
29
+ "epoch": 2.11,
30
+ "learning_rate": 7.841610738255034e-05,
31
+ "loss": 1.4122,
32
+ "step": 400
33
+ },
34
+ {
35
+ "epoch": 2.63,
36
+ "learning_rate": 7.787919463087249e-05,
37
+ "loss": 1.1658,
38
+ "step": 500
39
+ },
40
+ {
41
+ "epoch": 2.63,
42
+ "eval_las": 67.36639286076462,
43
+ "eval_loss": 1.6013022661209106,
44
+ "eval_runtime": 6.4078,
45
+ "eval_samples_per_second": 141.859,
46
+ "eval_steps_per_second": 17.791,
47
+ "eval_uas": 74.67802454112159,
48
+ "step": 500
49
+ },
50
+ {
51
+ "epoch": 3.16,
52
+ "learning_rate": 7.734228187919463e-05,
53
+ "loss": 1.0887,
54
+ "step": 600
55
+ },
56
+ {
57
+ "epoch": 3.68,
58
+ "learning_rate": 7.680536912751678e-05,
59
+ "loss": 0.9181,
60
+ "step": 700
61
+ },
62
+ {
63
+ "epoch": 4.21,
64
+ "learning_rate": 7.626845637583893e-05,
65
+ "loss": 0.8432,
66
+ "step": 800
67
+ },
68
+ {
69
+ "epoch": 4.74,
70
+ "learning_rate": 7.573154362416108e-05,
71
+ "loss": 0.7403,
72
+ "step": 900
73
+ },
74
+ {
75
+ "epoch": 5.26,
76
+ "learning_rate": 7.519463087248323e-05,
77
+ "loss": 0.6848,
78
+ "step": 1000
79
+ },
80
+ {
81
+ "epoch": 5.26,
82
+ "eval_las": 69.60416455396681,
83
+ "eval_loss": 1.8538918495178223,
84
+ "eval_runtime": 6.402,
85
+ "eval_samples_per_second": 141.987,
86
+ "eval_steps_per_second": 17.807,
87
+ "eval_uas": 76.92593719365853,
88
+ "step": 1000
89
+ },
90
+ {
91
+ "epoch": 5.79,
92
+ "learning_rate": 7.465771812080537e-05,
93
+ "loss": 0.6156,
94
+ "step": 1100
95
+ },
96
+ {
97
+ "epoch": 6.32,
98
+ "learning_rate": 7.412080536912752e-05,
99
+ "loss": 0.5488,
100
+ "step": 1200
101
+ },
102
+ {
103
+ "epoch": 6.84,
104
+ "learning_rate": 7.358389261744967e-05,
105
+ "loss": 0.5053,
106
+ "step": 1300
107
+ },
108
+ {
109
+ "epoch": 7.37,
110
+ "learning_rate": 7.304697986577182e-05,
111
+ "loss": 0.4459,
112
+ "step": 1400
113
+ },
114
+ {
115
+ "epoch": 7.89,
116
+ "learning_rate": 7.251006711409397e-05,
117
+ "loss": 0.4394,
118
+ "step": 1500
119
+ },
120
+ {
121
+ "epoch": 7.89,
122
+ "eval_las": 70.65882432478112,
123
+ "eval_loss": 1.9962742328643799,
124
+ "eval_runtime": 6.3907,
125
+ "eval_samples_per_second": 142.238,
126
+ "eval_steps_per_second": 17.838,
127
+ "eval_uas": 77.94679376669032,
128
+ "step": 1500
129
+ },
130
+ {
131
+ "epoch": 8.42,
132
+ "learning_rate": 7.197315436241612e-05,
133
+ "loss": 0.3671,
134
+ "step": 1600
135
+ },
136
+ {
137
+ "epoch": 8.95,
138
+ "learning_rate": 7.143624161073826e-05,
139
+ "loss": 0.3761,
140
+ "step": 1700
141
+ },
142
+ {
143
+ "epoch": 9.47,
144
+ "learning_rate": 7.089932885906041e-05,
145
+ "loss": 0.3076,
146
+ "step": 1800
147
+ },
148
+ {
149
+ "epoch": 10.0,
150
+ "learning_rate": 7.036241610738256e-05,
151
+ "loss": 0.3257,
152
+ "step": 1900
153
+ },
154
+ {
155
+ "epoch": 10.53,
156
+ "learning_rate": 6.982550335570471e-05,
157
+ "loss": 0.259,
158
+ "step": 2000
159
+ },
160
+ {
161
+ "epoch": 10.53,
162
+ "eval_las": 70.97657438393672,
163
+ "eval_loss": 2.430649995803833,
164
+ "eval_runtime": 6.3911,
165
+ "eval_samples_per_second": 142.229,
166
+ "eval_steps_per_second": 17.837,
167
+ "eval_uas": 78.3625730994152,
168
+ "step": 2000
169
+ },
170
+ {
171
+ "epoch": 11.05,
172
+ "learning_rate": 6.928859060402686e-05,
173
+ "loss": 0.2775,
174
+ "step": 2100
175
+ },
176
+ {
177
+ "epoch": 11.58,
178
+ "learning_rate": 6.8751677852349e-05,
179
+ "loss": 0.2325,
180
+ "step": 2200
181
+ },
182
+ {
183
+ "epoch": 12.11,
184
+ "learning_rate": 6.821476510067115e-05,
185
+ "loss": 0.2426,
186
+ "step": 2300
187
+ },
188
+ {
189
+ "epoch": 12.63,
190
+ "learning_rate": 6.76778523489933e-05,
191
+ "loss": 0.2116,
192
+ "step": 2400
193
+ },
194
+ {
195
+ "epoch": 13.16,
196
+ "learning_rate": 6.714093959731544e-05,
197
+ "loss": 0.2116,
198
+ "step": 2500
199
+ },
200
+ {
201
+ "epoch": 13.16,
202
+ "eval_las": 71.21319676841429,
203
+ "eval_loss": 2.8564882278442383,
204
+ "eval_runtime": 6.4121,
205
+ "eval_samples_per_second": 141.763,
206
+ "eval_steps_per_second": 17.779,
207
+ "eval_uas": 78.58905452455802,
208
+ "step": 2500
209
+ },
210
+ {
211
+ "epoch": 13.68,
212
+ "learning_rate": 6.660402684563758e-05,
213
+ "loss": 0.1948,
214
+ "step": 2600
215
+ },
216
+ {
217
+ "epoch": 14.21,
218
+ "learning_rate": 6.606711409395973e-05,
219
+ "loss": 0.1862,
220
+ "step": 2700
221
+ },
222
+ {
223
+ "epoch": 14.74,
224
+ "learning_rate": 6.553020134228188e-05,
225
+ "loss": 0.1754,
226
+ "step": 2800
227
+ },
228
+ {
229
+ "epoch": 15.26,
230
+ "learning_rate": 6.499328859060403e-05,
231
+ "loss": 0.1705,
232
+ "step": 2900
233
+ },
234
+ {
235
+ "epoch": 15.79,
236
+ "learning_rate": 6.445637583892618e-05,
237
+ "loss": 0.1618,
238
+ "step": 3000
239
+ },
240
+ {
241
+ "epoch": 15.79,
242
+ "eval_las": 71.65601865936517,
243
+ "eval_loss": 2.8342225551605225,
244
+ "eval_runtime": 6.4119,
245
+ "eval_samples_per_second": 141.768,
246
+ "eval_steps_per_second": 17.78,
247
+ "eval_uas": 78.93046682216138,
248
+ "step": 3000
249
+ },
250
+ {
251
+ "epoch": 16.32,
252
+ "learning_rate": 6.391946308724833e-05,
253
+ "loss": 0.1507,
254
+ "step": 3100
255
+ },
256
+ {
257
+ "epoch": 16.84,
258
+ "learning_rate": 6.338255033557047e-05,
259
+ "loss": 0.153,
260
+ "step": 3200
261
+ },
262
+ {
263
+ "epoch": 17.37,
264
+ "learning_rate": 6.284563758389262e-05,
265
+ "loss": 0.1401,
266
+ "step": 3300
267
+ },
268
+ {
269
+ "epoch": 17.89,
270
+ "learning_rate": 6.230872483221477e-05,
271
+ "loss": 0.1416,
272
+ "step": 3400
273
+ },
274
+ {
275
+ "epoch": 18.42,
276
+ "learning_rate": 6.177181208053692e-05,
277
+ "loss": 0.1273,
278
+ "step": 3500
279
+ },
280
+ {
281
+ "epoch": 18.42,
282
+ "eval_las": 71.52756650779165,
283
+ "eval_loss": 3.102910280227661,
284
+ "eval_runtime": 6.3959,
285
+ "eval_samples_per_second": 142.122,
286
+ "eval_steps_per_second": 17.824,
287
+ "eval_uas": 78.93384714193962,
288
+ "step": 3500
289
+ },
290
+ {
291
+ "epoch": 18.95,
292
+ "learning_rate": 6.123489932885907e-05,
293
+ "loss": 0.132,
294
+ "step": 3600
295
+ },
296
+ {
297
+ "epoch": 19.47,
298
+ "learning_rate": 6.069798657718121e-05,
299
+ "loss": 0.1188,
300
+ "step": 3700
301
+ },
302
+ {
303
+ "epoch": 20.0,
304
+ "learning_rate": 6.016107382550336e-05,
305
+ "loss": 0.1249,
306
+ "step": 3800
307
+ },
308
+ {
309
+ "epoch": 20.53,
310
+ "learning_rate": 5.9624161073825505e-05,
311
+ "loss": 0.1075,
312
+ "step": 3900
313
+ },
314
+ {
315
+ "epoch": 21.05,
316
+ "learning_rate": 5.9087248322147654e-05,
317
+ "loss": 0.1147,
318
+ "step": 4000
319
+ },
320
+ {
321
+ "epoch": 21.05,
322
+ "eval_las": 71.81151336916473,
323
+ "eval_loss": 3.3238770961761475,
324
+ "eval_runtime": 6.4049,
325
+ "eval_samples_per_second": 141.923,
326
+ "eval_steps_per_second": 17.799,
327
+ "eval_uas": 79.0048338572829,
328
+ "step": 4000
329
+ },
330
+ {
331
+ "epoch": 21.58,
332
+ "learning_rate": 5.85503355704698e-05,
333
+ "loss": 0.1037,
334
+ "step": 4100
335
+ },
336
+ {
337
+ "epoch": 22.11,
338
+ "learning_rate": 5.801342281879195e-05,
339
+ "loss": 0.1071,
340
+ "step": 4200
341
+ },
342
+ {
343
+ "epoch": 22.63,
344
+ "learning_rate": 5.74765100671141e-05,
345
+ "loss": 0.0993,
346
+ "step": 4300
347
+ },
348
+ {
349
+ "epoch": 23.16,
350
+ "learning_rate": 5.6939597315436247e-05,
351
+ "loss": 0.0986,
352
+ "step": 4400
353
+ },
354
+ {
355
+ "epoch": 23.68,
356
+ "learning_rate": 5.6402684563758395e-05,
357
+ "loss": 0.0916,
358
+ "step": 4500
359
+ },
360
+ {
361
+ "epoch": 23.68,
362
+ "eval_las": 71.7709495318257,
363
+ "eval_loss": 3.397751569747925,
364
+ "eval_runtime": 6.3879,
365
+ "eval_samples_per_second": 142.3,
366
+ "eval_steps_per_second": 17.846,
367
+ "eval_uas": 79.10286313085217,
368
+ "step": 4500
369
+ },
370
+ {
371
+ "epoch": 24.21,
372
+ "learning_rate": 5.586577181208054e-05,
373
+ "loss": 0.0936,
374
+ "step": 4600
375
+ },
376
+ {
377
+ "epoch": 24.74,
378
+ "learning_rate": 5.532885906040269e-05,
379
+ "loss": 0.0917,
380
+ "step": 4700
381
+ },
382
+ {
383
+ "epoch": 25.26,
384
+ "learning_rate": 5.479194630872483e-05,
385
+ "loss": 0.0881,
386
+ "step": 4800
387
+ },
388
+ {
389
+ "epoch": 25.79,
390
+ "learning_rate": 5.425503355704698e-05,
391
+ "loss": 0.0853,
392
+ "step": 4900
393
+ },
394
+ {
395
+ "epoch": 26.32,
396
+ "learning_rate": 5.371812080536913e-05,
397
+ "loss": 0.0839,
398
+ "step": 5000
399
+ },
400
+ {
401
+ "epoch": 26.32,
402
+ "eval_las": 72.05827671297705,
403
+ "eval_loss": 3.57505464553833,
404
+ "eval_runtime": 6.3979,
405
+ "eval_samples_per_second": 142.078,
406
+ "eval_steps_per_second": 17.818,
407
+ "eval_uas": 79.14680728796945,
408
+ "step": 5000
409
+ },
410
+ {
411
+ "epoch": 26.84,
412
+ "learning_rate": 5.318120805369128e-05,
413
+ "loss": 0.0815,
414
+ "step": 5100
415
+ },
416
+ {
417
+ "epoch": 27.37,
418
+ "learning_rate": 5.2644295302013426e-05,
419
+ "loss": 0.0777,
420
+ "step": 5200
421
+ },
422
+ {
423
+ "epoch": 27.89,
424
+ "learning_rate": 5.2107382550335574e-05,
425
+ "loss": 0.081,
426
+ "step": 5300
427
+ },
428
+ {
429
+ "epoch": 28.42,
430
+ "learning_rate": 5.157046979865772e-05,
431
+ "loss": 0.0723,
432
+ "step": 5400
433
+ },
434
+ {
435
+ "epoch": 28.95,
436
+ "learning_rate": 5.103355704697987e-05,
437
+ "loss": 0.0771,
438
+ "step": 5500
439
+ },
440
+ {
441
+ "epoch": 28.95,
442
+ "eval_las": 72.2408139810026,
443
+ "eval_loss": 3.521331787109375,
444
+ "eval_runtime": 6.426,
445
+ "eval_samples_per_second": 141.456,
446
+ "eval_steps_per_second": 17.74,
447
+ "eval_uas": 79.40371159111652,
448
+ "step": 5500
449
+ },
450
+ {
451
+ "epoch": 29.47,
452
+ "learning_rate": 5.049664429530202e-05,
453
+ "loss": 0.072,
454
+ "step": 5600
455
+ },
456
+ {
457
+ "epoch": 30.0,
458
+ "learning_rate": 4.995973154362417e-05,
459
+ "loss": 0.0745,
460
+ "step": 5700
461
+ },
462
+ {
463
+ "epoch": 30.53,
464
+ "learning_rate": 4.9422818791946315e-05,
465
+ "loss": 0.0681,
466
+ "step": 5800
467
+ },
468
+ {
469
+ "epoch": 31.05,
470
+ "learning_rate": 4.888590604026846e-05,
471
+ "loss": 0.069,
472
+ "step": 5900
473
+ },
474
+ {
475
+ "epoch": 31.58,
476
+ "learning_rate": 4.8348993288590605e-05,
477
+ "loss": 0.065,
478
+ "step": 6000
479
+ },
480
+ {
481
+ "epoch": 31.58,
482
+ "eval_las": 71.79461177027346,
483
+ "eval_loss": 3.818310022354126,
484
+ "eval_runtime": 6.3782,
485
+ "eval_samples_per_second": 142.517,
486
+ "eval_steps_per_second": 17.873,
487
+ "eval_uas": 79.04877801440017,
488
+ "step": 6000
489
+ },
490
+ {
491
+ "epoch": 32.11,
492
+ "learning_rate": 4.781208053691275e-05,
493
+ "loss": 0.066,
494
+ "step": 6100
495
+ },
496
+ {
497
+ "epoch": 32.63,
498
+ "learning_rate": 4.72751677852349e-05,
499
+ "loss": 0.0592,
500
+ "step": 6200
501
+ },
502
+ {
503
+ "epoch": 33.16,
504
+ "learning_rate": 4.673825503355705e-05,
505
+ "loss": 0.0637,
506
+ "step": 6300
507
+ },
508
+ {
509
+ "epoch": 33.68,
510
+ "learning_rate": 4.62013422818792e-05,
511
+ "loss": 0.0591,
512
+ "step": 6400
513
+ },
514
+ {
515
+ "epoch": 34.21,
516
+ "learning_rate": 4.5664429530201346e-05,
517
+ "loss": 0.0587,
518
+ "step": 6500
519
+ },
520
+ {
521
+ "epoch": 34.21,
522
+ "eval_las": 72.31180069634587,
523
+ "eval_loss": 3.807619333267212,
524
+ "eval_runtime": 6.3984,
525
+ "eval_samples_per_second": 142.068,
526
+ "eval_steps_per_second": 17.817,
527
+ "eval_uas": 79.51188182402055,
528
+ "step": 6500
529
+ },
530
+ {
531
+ "epoch": 34.74,
532
+ "learning_rate": 4.5127516778523494e-05,
533
+ "loss": 0.058,
534
+ "step": 6600
535
+ },
536
+ {
537
+ "epoch": 35.26,
538
+ "learning_rate": 4.459060402684564e-05,
539
+ "loss": 0.0572,
540
+ "step": 6700
541
+ },
542
+ {
543
+ "epoch": 35.79,
544
+ "learning_rate": 4.405369127516779e-05,
545
+ "loss": 0.0562,
546
+ "step": 6800
547
+ },
548
+ {
549
+ "epoch": 36.32,
550
+ "learning_rate": 4.351677852348994e-05,
551
+ "loss": 0.0547,
552
+ "step": 6900
553
+ },
554
+ {
555
+ "epoch": 36.84,
556
+ "learning_rate": 4.297986577181209e-05,
557
+ "loss": 0.0573,
558
+ "step": 7000
559
+ },
560
+ {
561
+ "epoch": 36.84,
562
+ "eval_las": 72.48419700503668,
563
+ "eval_loss": 3.9023566246032715,
564
+ "eval_runtime": 6.3817,
565
+ "eval_samples_per_second": 142.438,
566
+ "eval_steps_per_second": 17.864,
567
+ "eval_uas": 79.59638981847684,
568
+ "step": 7000
569
+ },
570
+ {
571
+ "epoch": 37.37,
572
+ "learning_rate": 4.244295302013423e-05,
573
+ "loss": 0.0522,
574
+ "step": 7100
575
+ },
576
+ {
577
+ "epoch": 37.89,
578
+ "learning_rate": 4.190604026845638e-05,
579
+ "loss": 0.0525,
580
+ "step": 7200
581
+ },
582
+ {
583
+ "epoch": 38.42,
584
+ "learning_rate": 4.1369127516778525e-05,
585
+ "loss": 0.0485,
586
+ "step": 7300
587
+ },
588
+ {
589
+ "epoch": 38.95,
590
+ "learning_rate": 4.0832214765100673e-05,
591
+ "loss": 0.0536,
592
+ "step": 7400
593
+ },
594
+ {
595
+ "epoch": 39.47,
596
+ "learning_rate": 4.029530201342282e-05,
597
+ "loss": 0.0503,
598
+ "step": 7500
599
+ },
600
+ {
601
+ "epoch": 39.47,
602
+ "eval_las": 72.57884595882771,
603
+ "eval_loss": 4.051434516906738,
604
+ "eval_runtime": 6.3781,
605
+ "eval_samples_per_second": 142.52,
606
+ "eval_steps_per_second": 17.874,
607
+ "eval_uas": 79.66399621404184,
608
+ "step": 7500
609
+ },
610
+ {
611
+ "epoch": 40.0,
612
+ "learning_rate": 3.975838926174497e-05,
613
+ "loss": 0.0488,
614
+ "step": 7600
615
+ },
616
+ {
617
+ "epoch": 40.53,
618
+ "learning_rate": 3.922147651006712e-05,
619
+ "loss": 0.0459,
620
+ "step": 7700
621
+ },
622
+ {
623
+ "epoch": 41.05,
624
+ "learning_rate": 3.8684563758389266e-05,
625
+ "loss": 0.0481,
626
+ "step": 7800
627
+ },
628
+ {
629
+ "epoch": 41.58,
630
+ "learning_rate": 3.814765100671141e-05,
631
+ "loss": 0.0464,
632
+ "step": 7900
633
+ },
634
+ {
635
+ "epoch": 42.11,
636
+ "learning_rate": 3.7610738255033556e-05,
637
+ "loss": 0.0476,
638
+ "step": 8000
639
+ },
640
+ {
641
+ "epoch": 42.11,
642
+ "eval_las": 72.74786194774026,
643
+ "eval_loss": 4.076736927032471,
644
+ "eval_runtime": 6.3895,
645
+ "eval_samples_per_second": 142.265,
646
+ "eval_steps_per_second": 17.842,
647
+ "eval_uas": 79.88033667984992,
648
+ "step": 8000
649
+ },
650
+ {
651
+ "epoch": 42.63,
652
+ "learning_rate": 3.7073825503355704e-05,
653
+ "loss": 0.0456,
654
+ "step": 8100
655
+ },
656
+ {
657
+ "epoch": 43.16,
658
+ "learning_rate": 3.653691275167785e-05,
659
+ "loss": 0.0464,
660
+ "step": 8200
661
+ },
662
+ {
663
+ "epoch": 43.68,
664
+ "learning_rate": 3.6e-05,
665
+ "loss": 0.0429,
666
+ "step": 8300
667
+ },
668
+ {
669
+ "epoch": 44.21,
670
+ "learning_rate": 3.546308724832215e-05,
671
+ "loss": 0.0442,
672
+ "step": 8400
673
+ },
674
+ {
675
+ "epoch": 44.74,
676
+ "learning_rate": 3.49261744966443e-05,
677
+ "loss": 0.0427,
678
+ "step": 8500
679
+ },
680
+ {
681
+ "epoch": 44.74,
682
+ "eval_las": 72.79180610485751,
683
+ "eval_loss": 4.197169303894043,
684
+ "eval_runtime": 6.3929,
685
+ "eval_samples_per_second": 142.188,
686
+ "eval_steps_per_second": 17.832,
687
+ "eval_uas": 79.76540580738939,
688
+ "step": 8500
689
+ },
690
+ {
691
+ "epoch": 45.26,
692
+ "learning_rate": 3.4389261744966446e-05,
693
+ "loss": 0.0405,
694
+ "step": 8600
695
+ },
696
+ {
697
+ "epoch": 45.79,
698
+ "learning_rate": 3.3852348993288594e-05,
699
+ "loss": 0.0416,
700
+ "step": 8700
701
+ },
702
+ {
703
+ "epoch": 46.32,
704
+ "learning_rate": 3.331543624161074e-05,
705
+ "loss": 0.0402,
706
+ "step": 8800
707
+ },
708
+ {
709
+ "epoch": 46.84,
710
+ "learning_rate": 3.277852348993289e-05,
711
+ "loss": 0.0389,
712
+ "step": 8900
713
+ },
714
+ {
715
+ "epoch": 47.37,
716
+ "learning_rate": 3.224161073825504e-05,
717
+ "loss": 0.0415,
718
+ "step": 9000
719
+ },
720
+ {
721
+ "epoch": 47.37,
722
+ "eval_las": 72.96758273332658,
723
+ "eval_loss": 4.470444202423096,
724
+ "eval_runtime": 6.3862,
725
+ "eval_samples_per_second": 142.338,
726
+ "eval_steps_per_second": 17.851,
727
+ "eval_uas": 79.90737923807593,
728
+ "step": 9000
729
+ },
730
+ {
731
+ "epoch": 47.89,
732
+ "learning_rate": 3.170469798657718e-05,
733
+ "loss": 0.0402,
734
+ "step": 9100
735
+ },
736
+ {
737
+ "epoch": 48.42,
738
+ "learning_rate": 3.116778523489933e-05,
739
+ "loss": 0.036,
740
+ "step": 9200
741
+ },
742
+ {
743
+ "epoch": 48.95,
744
+ "learning_rate": 3.0630872483221477e-05,
745
+ "loss": 0.0391,
746
+ "step": 9300
747
+ },
748
+ {
749
+ "epoch": 49.47,
750
+ "learning_rate": 3.0093959731543628e-05,
751
+ "loss": 0.0351,
752
+ "step": 9400
753
+ },
754
+ {
755
+ "epoch": 50.0,
756
+ "learning_rate": 2.9557046979865776e-05,
757
+ "loss": 0.0389,
758
+ "step": 9500
759
+ },
760
+ {
761
+ "epoch": 50.0,
762
+ "eval_las": 72.97434337288308,
763
+ "eval_loss": 4.371275901794434,
764
+ "eval_runtime": 6.4275,
765
+ "eval_samples_per_second": 141.424,
766
+ "eval_steps_per_second": 17.736,
767
+ "eval_uas": 80.02231011053645,
768
+ "step": 9500
769
+ },
770
+ {
771
+ "epoch": 50.53,
772
+ "learning_rate": 2.9020134228187925e-05,
773
+ "loss": 0.036,
774
+ "step": 9600
775
+ },
776
+ {
777
+ "epoch": 51.05,
778
+ "learning_rate": 2.848322147651007e-05,
779
+ "loss": 0.0365,
780
+ "step": 9700
781
+ },
782
+ {
783
+ "epoch": 51.58,
784
+ "learning_rate": 2.7946308724832218e-05,
785
+ "loss": 0.0348,
786
+ "step": 9800
787
+ },
788
+ {
789
+ "epoch": 52.11,
790
+ "learning_rate": 2.7409395973154366e-05,
791
+ "loss": 0.0338,
792
+ "step": 9900
793
+ },
794
+ {
795
+ "epoch": 52.63,
796
+ "learning_rate": 2.6872483221476514e-05,
797
+ "loss": 0.0339,
798
+ "step": 10000
799
+ },
800
+ {
801
+ "epoch": 52.63,
802
+ "eval_las": 73.21772639691714,
803
+ "eval_loss": 4.721055507659912,
804
+ "eval_runtime": 6.3848,
805
+ "eval_samples_per_second": 142.369,
806
+ "eval_steps_per_second": 17.855,
807
+ "eval_uas": 80.05273298854073,
808
+ "step": 10000
809
+ },
810
+ {
811
+ "epoch": 53.16,
812
+ "learning_rate": 2.6335570469798663e-05,
813
+ "loss": 0.0336,
814
+ "step": 10100
815
+ },
816
+ {
817
+ "epoch": 53.68,
818
+ "learning_rate": 2.5798657718120804e-05,
819
+ "loss": 0.0338,
820
+ "step": 10200
821
+ },
822
+ {
823
+ "epoch": 54.21,
824
+ "learning_rate": 2.5261744966442952e-05,
825
+ "loss": 0.0312,
826
+ "step": 10300
827
+ },
828
+ {
829
+ "epoch": 54.74,
830
+ "learning_rate": 2.47248322147651e-05,
831
+ "loss": 0.0329,
832
+ "step": 10400
833
+ },
834
+ {
835
+ "epoch": 55.26,
836
+ "learning_rate": 2.4187919463087252e-05,
837
+ "loss": 0.0354,
838
+ "step": 10500
839
+ },
840
+ {
841
+ "epoch": 55.26,
842
+ "eval_las": 72.97096305310482,
843
+ "eval_loss": 4.761220455169678,
844
+ "eval_runtime": 6.3891,
845
+ "eval_samples_per_second": 142.275,
846
+ "eval_steps_per_second": 17.843,
847
+ "eval_uas": 79.91075955785418,
848
+ "step": 10500
849
+ },
850
+ {
851
+ "epoch": 55.79,
852
+ "learning_rate": 2.36510067114094e-05,
853
+ "loss": 0.0312,
854
+ "step": 10600
855
+ },
856
+ {
857
+ "epoch": 56.32,
858
+ "learning_rate": 2.311409395973155e-05,
859
+ "loss": 0.0322,
860
+ "step": 10700
861
+ },
862
+ {
863
+ "epoch": 56.84,
864
+ "learning_rate": 2.257718120805369e-05,
865
+ "loss": 0.0298,
866
+ "step": 10800
867
+ },
868
+ {
869
+ "epoch": 57.37,
870
+ "learning_rate": 2.204026845637584e-05,
871
+ "loss": 0.0304,
872
+ "step": 10900
873
+ },
874
+ {
875
+ "epoch": 57.89,
876
+ "learning_rate": 2.1503355704697987e-05,
877
+ "loss": 0.0304,
878
+ "step": 11000
879
+ },
880
+ {
881
+ "epoch": 57.89,
882
+ "eval_las": 73.20420511780415,
883
+ "eval_loss": 4.784778594970703,
884
+ "eval_runtime": 6.3775,
885
+ "eval_samples_per_second": 142.533,
886
+ "eval_steps_per_second": 17.875,
887
+ "eval_uas": 80.04935266876247,
888
+ "step": 11000
889
+ },
890
+ {
891
+ "epoch": 58.42,
892
+ "learning_rate": 2.0966442953020135e-05,
893
+ "loss": 0.0278,
894
+ "step": 11100
895
+ },
896
+ {
897
+ "epoch": 58.95,
898
+ "learning_rate": 2.0429530201342283e-05,
899
+ "loss": 0.0314,
900
+ "step": 11200
901
+ },
902
+ {
903
+ "epoch": 59.47,
904
+ "learning_rate": 1.989261744966443e-05,
905
+ "loss": 0.029,
906
+ "step": 11300
907
+ },
908
+ {
909
+ "epoch": 60.0,
910
+ "learning_rate": 1.935570469798658e-05,
911
+ "loss": 0.0278,
912
+ "step": 11400
913
+ },
914
+ {
915
+ "epoch": 60.53,
916
+ "learning_rate": 1.8818791946308724e-05,
917
+ "loss": 0.0273,
918
+ "step": 11500
919
+ },
920
+ {
921
+ "epoch": 60.53,
922
+ "eval_las": 73.17716255957814,
923
+ "eval_loss": 5.09051513671875,
924
+ "eval_runtime": 6.4269,
925
+ "eval_samples_per_second": 141.436,
926
+ "eval_steps_per_second": 17.738,
927
+ "eval_uas": 80.10005746543622,
928
+ "step": 11500
929
+ },
930
+ {
931
+ "epoch": 61.05,
932
+ "learning_rate": 1.8281879194630873e-05,
933
+ "loss": 0.0296,
934
+ "step": 11600
935
+ },
936
+ {
937
+ "epoch": 61.58,
938
+ "learning_rate": 1.774496644295302e-05,
939
+ "loss": 0.0281,
940
+ "step": 11700
941
+ },
942
+ {
943
+ "epoch": 62.11,
944
+ "learning_rate": 1.720805369127517e-05,
945
+ "loss": 0.0259,
946
+ "step": 11800
947
+ },
948
+ {
949
+ "epoch": 62.63,
950
+ "learning_rate": 1.6671140939597317e-05,
951
+ "loss": 0.0264,
952
+ "step": 11900
953
+ },
954
+ {
955
+ "epoch": 63.16,
956
+ "learning_rate": 1.6134228187919466e-05,
957
+ "loss": 0.0273,
958
+ "step": 12000
959
+ },
960
+ {
961
+ "epoch": 63.16,
962
+ "eval_las": 73.50505357806848,
963
+ "eval_loss": 5.068530559539795,
964
+ "eval_runtime": 6.3815,
965
+ "eval_samples_per_second": 142.442,
966
+ "eval_steps_per_second": 17.864,
967
+ "eval_uas": 80.3840043268093,
968
+ "step": 12000
969
+ },
970
+ {
971
+ "epoch": 63.68,
972
+ "learning_rate": 1.559731543624161e-05,
973
+ "loss": 0.026,
974
+ "step": 12100
975
+ },
976
+ {
977
+ "epoch": 64.21,
978
+ "learning_rate": 1.506040268456376e-05,
979
+ "loss": 0.0269,
980
+ "step": 12200
981
+ },
982
+ {
983
+ "epoch": 64.74,
984
+ "learning_rate": 1.4523489932885909e-05,
985
+ "loss": 0.0243,
986
+ "step": 12300
987
+ },
988
+ {
989
+ "epoch": 65.26,
990
+ "learning_rate": 1.3986577181208053e-05,
991
+ "loss": 0.0266,
992
+ "step": 12400
993
+ },
994
+ {
995
+ "epoch": 65.79,
996
+ "learning_rate": 1.3449664429530202e-05,
997
+ "loss": 0.0256,
998
+ "step": 12500
999
+ },
1000
+ {
1001
+ "epoch": 65.79,
1002
+ "eval_las": 73.54561741540749,
1003
+ "eval_loss": 5.208915710449219,
1004
+ "eval_runtime": 6.3736,
1005
+ "eval_samples_per_second": 142.619,
1006
+ "eval_steps_per_second": 17.886,
1007
+ "eval_uas": 80.3231585708008,
1008
+ "step": 12500
1009
+ },
1010
+ {
1011
+ "epoch": 66.32,
1012
+ "learning_rate": 1.2912751677852352e-05,
1013
+ "loss": 0.0245,
1014
+ "step": 12600
1015
+ },
1016
+ {
1017
+ "epoch": 66.84,
1018
+ "learning_rate": 1.2375838926174497e-05,
1019
+ "loss": 0.0265,
1020
+ "step": 12700
1021
+ },
1022
+ {
1023
+ "epoch": 67.37,
1024
+ "learning_rate": 1.1838926174496645e-05,
1025
+ "loss": 0.026,
1026
+ "step": 12800
1027
+ },
1028
+ {
1029
+ "epoch": 67.89,
1030
+ "learning_rate": 1.1302013422818795e-05,
1031
+ "loss": 0.0232,
1032
+ "step": 12900
1033
+ },
1034
+ {
1035
+ "epoch": 68.42,
1036
+ "learning_rate": 1.076510067114094e-05,
1037
+ "loss": 0.0231,
1038
+ "step": 13000
1039
+ },
1040
+ {
1041
+ "epoch": 68.42,
1042
+ "eval_las": 73.63350572964202,
1043
+ "eval_loss": 5.212841033935547,
1044
+ "eval_runtime": 6.4593,
1045
+ "eval_samples_per_second": 140.728,
1046
+ "eval_steps_per_second": 17.649,
1047
+ "eval_uas": 80.35020112902681,
1048
+ "step": 13000
1049
+ },
1050
+ {
1051
+ "epoch": 68.95,
1052
+ "learning_rate": 1.0228187919463088e-05,
1053
+ "loss": 0.0242,
1054
+ "step": 13100
1055
+ },
1056
+ {
1057
+ "epoch": 69.47,
1058
+ "learning_rate": 9.691275167785236e-06,
1059
+ "loss": 0.0256,
1060
+ "step": 13200
1061
+ },
1062
+ {
1063
+ "epoch": 70.0,
1064
+ "learning_rate": 9.154362416107383e-06,
1065
+ "loss": 0.0249,
1066
+ "step": 13300
1067
+ },
1068
+ {
1069
+ "epoch": 70.53,
1070
+ "learning_rate": 8.617449664429531e-06,
1071
+ "loss": 0.0236,
1072
+ "step": 13400
1073
+ },
1074
+ {
1075
+ "epoch": 71.05,
1076
+ "learning_rate": 8.080536912751679e-06,
1077
+ "loss": 0.0237,
1078
+ "step": 13500
1079
+ },
1080
+ {
1081
+ "epoch": 71.05,
1082
+ "eval_las": 73.65716796808978,
1083
+ "eval_loss": 5.283775806427002,
1084
+ "eval_runtime": 6.3678,
1085
+ "eval_samples_per_second": 142.751,
1086
+ "eval_steps_per_second": 17.903,
1087
+ "eval_uas": 80.42794848392658,
1088
+ "step": 13500
1089
+ },
1090
+ {
1091
+ "epoch": 71.58,
1092
+ "learning_rate": 7.543624161073826e-06,
1093
+ "loss": 0.0235,
1094
+ "step": 13600
1095
+ },
1096
+ {
1097
+ "epoch": 72.11,
1098
+ "learning_rate": 7.006711409395974e-06,
1099
+ "loss": 0.0221,
1100
+ "step": 13700
1101
+ },
1102
+ {
1103
+ "epoch": 72.63,
1104
+ "learning_rate": 6.469798657718121e-06,
1105
+ "loss": 0.0223,
1106
+ "step": 13800
1107
+ },
1108
+ {
1109
+ "epoch": 73.16,
1110
+ "learning_rate": 5.932885906040269e-06,
1111
+ "loss": 0.0227,
1112
+ "step": 13900
1113
+ },
1114
+ {
1115
+ "epoch": 73.68,
1116
+ "learning_rate": 5.395973154362416e-06,
1117
+ "loss": 0.0205,
1118
+ "step": 14000
1119
+ },
1120
+ {
1121
+ "epoch": 73.68,
1122
+ "eval_las": 73.76871852077207,
1123
+ "eval_loss": 5.542043209075928,
1124
+ "eval_runtime": 6.3741,
1125
+ "eval_samples_per_second": 142.609,
1126
+ "eval_steps_per_second": 17.885,
1127
+ "eval_uas": 80.4887942399351,
1128
+ "step": 14000
1129
+ },
1130
+ {
1131
+ "epoch": 74.21,
1132
+ "learning_rate": 4.859060402684564e-06,
1133
+ "loss": 0.023,
1134
+ "step": 14100
1135
+ },
1136
+ {
1137
+ "epoch": 74.74,
1138
+ "learning_rate": 4.322147651006712e-06,
1139
+ "loss": 0.0227,
1140
+ "step": 14200
1141
+ },
1142
+ {
1143
+ "epoch": 75.26,
1144
+ "learning_rate": 3.7852348993288595e-06,
1145
+ "loss": 0.0222,
1146
+ "step": 14300
1147
+ },
1148
+ {
1149
+ "epoch": 75.79,
1150
+ "learning_rate": 3.248322147651007e-06,
1151
+ "loss": 0.023,
1152
+ "step": 14400
1153
+ },
1154
+ {
1155
+ "epoch": 76.32,
1156
+ "learning_rate": 2.716778523489933e-06,
1157
+ "loss": 0.0221,
1158
+ "step": 14500
1159
+ },
1160
+ {
1161
+ "epoch": 76.32,
1162
+ "eval_las": 73.7112530845418,
1163
+ "eval_loss": 5.52190637588501,
1164
+ "eval_runtime": 6.3515,
1165
+ "eval_samples_per_second": 143.116,
1166
+ "eval_steps_per_second": 17.949,
1167
+ "eval_uas": 80.37048304769631,
1168
+ "step": 14500
1169
+ },
1170
+ {
1171
+ "epoch": 76.84,
1172
+ "learning_rate": 2.1798657718120807e-06,
1173
+ "loss": 0.0212,
1174
+ "step": 14600
1175
+ },
1176
+ {
1177
+ "epoch": 77.37,
1178
+ "learning_rate": 1.6429530201342283e-06,
1179
+ "loss": 0.021,
1180
+ "step": 14700
1181
+ },
1182
+ {
1183
+ "epoch": 77.89,
1184
+ "learning_rate": 1.1060402684563759e-06,
1185
+ "loss": 0.0226,
1186
+ "step": 14800
1187
+ },
1188
+ {
1189
+ "epoch": 78.42,
1190
+ "learning_rate": 5.691275167785235e-07,
1191
+ "loss": 0.0215,
1192
+ "step": 14900
1193
+ },
1194
+ {
1195
+ "epoch": 78.95,
1196
+ "learning_rate": 3.2214765100671145e-08,
1197
+ "loss": 0.0215,
1198
+ "step": 15000
1199
+ },
1200
+ {
1201
+ "epoch": 78.95,
1202
+ "eval_las": 73.77547916032857,
1203
+ "eval_loss": 5.533295154571533,
1204
+ "eval_runtime": 6.3578,
1205
+ "eval_samples_per_second": 142.973,
1206
+ "eval_steps_per_second": 17.931,
1207
+ "eval_uas": 80.46175168170909,
1208
+ "step": 15000
1209
+ },
1210
+ {
1211
+ "epoch": 78.95,
1212
+ "step": 15000,
1213
+ "total_flos": 9.574431893701632e+16,
1214
+ "train_loss": 0.1834182323773702,
1215
+ "train_runtime": 7377.9639,
1216
+ "train_samples_per_second": 65.059,
1217
+ "train_steps_per_second": 2.033
1218
+ }
1219
+ ],
1220
+ "max_steps": 15000,
1221
+ "num_train_epochs": 79,
1222
+ "total_flos": 9.574431893701632e+16,
1223
+ "trial_name": null,
1224
+ "trial_params": null
1225
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:264ec7b9f23f702d0029ea408bc8a3e6ff8a796ab3d9d09a00b1f93db72156f4
3
+ size 3387